Hi there, I’m Joseph.

I’m an Alignment Research Engineer. I train and analyse AI systems to work out how we can make them aligned with human values.

I studied computational biology and statistics and worked as a Data Scientist in a nascent start-up.

Humanity deserves to be safe.

In the last 10 years we have seen startling advances in the field of artificial intelligence. Whereas before we used machine learning to find statistical associations useful to market products, now we making AI which talk to us, create amazing works of art and are competitive at games like Diplomacy. This article provides a great overview of just how rapid advances are these data: Visualizing the deep learning revolution by Richard Ngo.

While these advances sound impressive, they raise serious questions about what the world will look like in the near future. The most important of these questions is: What happens to humanity if we create highly intelligent AI systems, smarter even, than ourselves? Many thinkers have pondered this question and the results are not good. The short story is, by default, AI will not be beneficial to humanity and we may even be bringing about our own extinction or permanent loss of control over our own destiny.

I’m trying to help.

It’s not clear yet what we can do to ensure that future AI will protect and nurture humanity. However, there are communities of people exploring how we might do this. You can find information about my current projects on the projects page.

About this site

This cite is generated with Hugo, using the Recluse theme created by Will Fondrie.